https://buxizhou.net不系舟记Halo v2.24.2zh-cnWed, 10 Jun 2026 01:15:31 GMThttps://buxizhou.net/archives/The%20Time%20Dimension%20in%20Machine%20Learning%20and%20Reinforcement%20Learning强化学习为什么更适合解决动态问题？它不需要预设的标准答案，而是靠智能体在环境里的摸爬滚打，自己找出一套收益最大化的策略。]]>/archives/The%20Time%20Dimension%20in%20Machine%20Learning%20and%20Reinforcement%20LearningbuxizhouRLFri, 29 May 2026 13:56:29 GMT